Conventional optimal torque control （OTC） strategies suffer from slow rotor speed regulation， which limits wind energy capture efficiency. To address this issue， a novel OTC strategy for wind turbines based on deep deterministic policy gradient （DDPG） and inertial compensation control （ICC） is proposed. Using the DDPG algorithm， the gain coefficient is updated online in real time according to wind speed and rotor speed. To further enhance rotor speed tracking dynamics， an inertial compensation term is introduced based on the ICC principle. To eliminate the dependence of the compensation term on aerodynamic torque sensors or mechanism-based modeling， a broad learning system （BLS） is employed to estimate the aerodynamic torque. Simulations conducted on the OpenFAST demonstrate that the proposed method significantly improves rotor speed tracking dynamics， thereby enhancing wind energy capture efficiency.
Driven by the “dual-carbon” goals， microgrid operation must simultaneously ensure economic efficiency and low-carbon performance. To address the uncertainties in photovoltaic （PV） generation and load forecasting， as well as the significant time-series variability of carbon emissions， a two-stage optimal dispatch model for microgrids under a carbon trading mechanism with time-series constraints is developed. Digital models of gas turbines， energy storage devices， demand response， and the carbon trading mechanism are established and incorporated into a unified optimization framework. To further enhance carbon reduction performance under the carbon trading mechanism， a high-carbon-pressure period identification method based on a quantile threshold is proposed， and a corresponding time-series constraint strategy is designed. Robust optimization is adopted to handle uncertainties in PV output and load forecasting. Case studies demonstrate that the proposed method can further reduce system carbon emissions while maintaining nearly unchanged total costs. Moreover， when source–load uncertainty increases， the model retains satisfactory dispatch stability and economic performance.
The circulating water system of two ultra-supercritical generating units in a power plant consists of six pumps， including fixed-speed pumps， dual-speed pumps， and variable-frequency pumps， resulting in relatively complex operating modes. To address this issue， an optimal operation method for the circulating water system is proposed. A refined steam turbine model coupled with the condenser-side system is developed using EBSILON software. Based on model simulations and function fitting， the key characteristic curves required for circulating water system optimization and condenser-side performance evaluation are determined. A hybrid optimization method combining basin hopping and the sequential least squares quadratic programming （SLSQP） is proposed. With maximum unit profit power as the objective， optimization studies are conducted for both dual-speed pump and variable-frequency pump operating modes. The optimal combinations of circulating water pumps under various operating conditions and the corresponding net power gains are obtained， demonstrating the effectiveness of the proposed method. The method provides an effective modeling tool and decision-making approach for energy-efficient optimal operation of complex circulating water systems in similar units.
As the integration between inter-provincial and intra-provincial electricity spot markets in China becomes increasingly tight， identifying risk spillovers across markets has become critical to ensuring stable market operations. This study investigates the mechanisms of price risk spillover and establishes an analytical framework based on the Diebold-Yilmaz （DY） spillover index and the Conditional Value at Risk （CoVaR） model. Risk spillover characteristics are examined from two dimensions： normal price fluctuations and extreme market conditions. Case studies are conducted to calculate and analyze the price risk spillover effects using historical transaction price data for 2024 from the inter-provincial and intra-provincial spot markets in Shanxi Province， the intra-provincial market in Shandong Province， and the inter-provincial market in Zhejiang Province， respectively. The results indicate that price risk spillovers among the sampled markets in 2024 exhibit significant asymmetry： strong bidirectional spillover effects are observed among intra-provincial markets， while limited spillovers exist between inter-provincial and intra-provincial markets. The risk spillover effect among inter-provincial markets is found to be the weakest.
To address the increased difficulty of voltage regulation in weak power grids with high renewable energy penetration， this paper proposes an optimal scheduling strategy for energy storage systems （ESSs） to participate in voltage regulation through coordinated charging and discharging， together with a corresponding ancillary service compensation mechanism. First， a three-stage day-ahead coordinated scheduling framework for a provincial power grid is established， consisting of day-ahead spot market clearing， day-ahead reactive power scheduling， and an ESS-based voltage regulation stage. Next， an optimal scheduling model centered on adjusting the day-ahead peak-shaving schedule of ESSs while compensating for the resulting loss of peak-shaving revenue is developed for the voltage regulation stage to support system voltage by regulating the active power output of ESSs. Subsequently， a compensation and cost allocation scheme for voltage regulation ancillary services is designed to align with the proposed scheduling model， ensuring voltage security while maintaining the reasonable economic returns of ESSs as independent market participants. Finally， simulations conducted on a modified IEEE 39-bus system and a power grid in Xizang verify the effectiveness of the proposed method.
To improve energy utilization efficiency and operational flexibility in multiple rural integrated energy systems （IESs）， a distributionally robust optimization strategy is proposed for coordinated operation considering electricity interaction among systems. First， to address uncertainties in renewable energy generation and load demand， a distributionally robust optimization （DRO） model based on the Wasserstein distance is introduced， which is further reformulated into a tractable deterministic equivalent model using Lipschitz function theory. Second， a cooperative operation model for electricity sharing among multiple rural IESs is developed based on Nash bargaining theory， aiming to minimize the total cost of coordinated system operation. To handle disparities in electricity interaction among participating entities， a profit allocation method based on an asymmetric Nash bargaining mechanism is proposed， and the alternating direction method of multipliers （ADMM） is adopted for solution， enabling fair distribution of cooperative benefits. Simulation results demonstrate that the proposed strategy effectively reduces system operating costs while achieving fair and efficient cooperative operation among rural integrated energy systems.
In the actual operation of inter-provincial electricity markets， it is essential to reasonably prevent the operational security risks of large power grids while maximizing the utilization efficiency of inter-provincial transmission corridors. To this end， a bi-level game-theoretic bidding model for provincial power grids based on the maximum trading capacity of interprovincial and interregional transmission corridors is established. First， a interprovincial spot trading network model considering physical grid operation constraints is constructed， based on which an optimization method for the maximum available capacity of transmission corridors is proposed. On this basis， a bi-level game-theoretic bidding model for provincial grid power procurement based on the maximum trading capacity of transmission corridors is formulated. Simulation results demonstrate that the proposed optimization method effectively coordinates grid security and market efficiency， significantly enhances the trading capacity of interprovincial corridors， and avoids interprovincial transit power flows and intraprovincial interface congestion. Furthermore， the bi-level game-theoretic bidding model yields the optimal power purchase bidding strategy under Nash equilibrium， providing a decision-making reference for market participants.
Grid maintenance， operating mode adjustments， and equipment commissioning may lead to dynamic changes in grid topology during long-term electricity spot market simulations. Rebuilding a complete power grid model whenever the grid topology changes significantly reduces the computational efficiency of long-term market simulations. To address this issue， this paper proposes a unified representation framework based on fast shift factor updating methods for three typical scenarios， with generalized branch closing serving as the basic operational unit. A generalized branch closing representation model is established to equivalently transform various complex topology-change scenarios that would otherwise require separate treatment into combinations of single branch-closing operations. On this basis， a long-term simulation workflow considering dynamic changes in grid topology is developed， enabling effective integration of time-varying topological characteristics with long-timescale market simulation. Numerical case studies demonstrate that the proposed method delivers efficient and stable computational performance across power grids of different scales and under different extents of topology change. Moreover， its computational efficiency advantage becomes increasingly pronounced as the grid scale increases.
With the increasing penetration of renewable generation， transmission congestion has become more frequent， rendering conventional system-wide structural market power indicators inadequate for detecting localized market power arising from network congestion. To address this challenge， this paper proposes a market power monitoring and mitigation mechanism for congestion-prone power systems. First， a localized structural market power monitoring is performed by conducting structural pivotal supplier （SPS） tests for generators associated with the transmission interface under evaluation， enabling the identification of market power risks arising from physical network constraints. Subsequently， a behavioral market power monitoring system is developed to detect strategic bidding behavior by SPS units. Finally， an ex-post excess profit testing and clawback mechanism is introduced to recover excessive revenues and return them to the market. Validation using market data and case studies from the Zhejiang power grid demonstrates that the proposed approach effectively identifies strategic behaviors， including high-price bidding and short-term bid price inflation， during periods of network congestion and implements corresponding mitigation measures. The results further confirm its strong practical applicability.
To address the dispatch challenges and node voltage violations caused by renewable energy output fluctuations and electric vehicle load variations in multi-microgrid distribution systems， a joint model–data-driven economic dispatch and voltage control method is proposed. First， the multi-microgrid distribution system is transformed as a Stackelberg game model， in which dynamic electricity pricing is designed to improve power interactions and benefit coordination among multiple agents. An embedded active price-updating loop is introduced to enhance the nonlinear adaptability of the day-ahead dispatch model while improving computational efficiency. Subsequently， deep reinforcement learning （RL） is employed to track day-ahead dispatch commands and implement real-time decentralized voltage control. This enables the distribution system to maintain autonomous reactive power regulation capability in the presence of unmodeled power disturbances while ensuring multi-agent economic dispatch. Simulation results demonstrate that the proposed method mitigates local power imbalance and dispatch burdens caused by excessive economically driven operation of energy storage systems under peak–valley price incentives， improves power interactions among different entities in the distribution system， and achieves active voltage control under source–load uncertainty conditions.
With the increasing penetration of distributed generation （DG） in power systems， the safety issues triggered by unintentional islanding have become increasingly prominent， posing a severe threat to grid operation. This paper presents a comprehensive review of the current mainstream islanding detection techniques and classifies them into three categories according to their detection principles and implementation locations： remote detection methods， local passive detection methods， and local active detection methods. The operating principles， technical characteristics， and latest research progress of each detection method are analyzed， and their respective advantages and disadvantages are critically evaluated. To address major challenges associated with existing techniques， including large non-detection zones （NDZs） and adverse impacts on power quality， several future research directions are identified， including multi-criterion information fusion， intelligent algorithm enhancement， improved signal-injection-based protection schemes， coordinated operation of protection and reclosing functions， and the refinement and validation of simulation models. The review provides key technical support for the construction of modern power systems with high penetrations of distributed generation.
In high-power DC conversion applications， CLLLC resonant converters with an ISOP （input-series output-parallel） configuration are widely used due to their low device voltage stress and high power density. However， affected by parameter dispersion and the negative impedance characteristic of constant power loads， input voltage imbalance remains a critical issue. To address this， an adaptive virtual impedance-based voltage balancing strategy with dynamic error perception is proposed. An 11th-order small-signal model of the CLLLC converter is established using the extended describing function （EDF） method， the physical significance of the state-space coefficient matrix is clarified， and a simplified expression of input impedance is derived to reveal the mechanism of voltage imbalance. A virtual resistance Rv and virtual inductance Lv are designed to vary nonlinearly with voltage error. Under large transient deviations， Rv is increased and the leading compensation effect of Lv is enhanced to accelerate convergence； under small steady-state deviations， Rv is reduced to decrease equivalent voltage drop and ripple coupling， while Lv is decreased to suppress differential noise injection and frequency-tuning oscillations. Experimental results demonstrate that the proposed strategy reshapes the equivalent input impedance of submodules without additional hardware and significantly improves voltage balancing performance. Compared with conventional methods， it achieves smaller steady-state voltage deviation and stronger robustness against parameter mismatch and operating condition disturbances， providing a theoretical basis for reliable operation of high-power systems.
Existing frequency support control strategies for wind turbines suffer from response delays， may trigger secondary frequency dips， and often involve coarse energy management. To address these issues， an improved frequency support strategy is proposed that incorporates an optimal frequency trajectory into model predictive control （MPC）. First， a prediction equation is established based on the system frequency response model， in which the frequency deviation is used as the state variable. An extended state observer is introduced to estimate grid active-power disturbances in real time and provide feedforward compensation. Second， using a predefined optimal frequency trajectory as the tracking target， an MPC rolling optimizer is designed to compute the optimal supplementary power command while satisfying safety constraints on turbine rotor speed and output power. A complete control logic incorporating mode switching and smooth exit is also developed. Finally， simulation studies on the 3-machine 9-bus and 10-machine 39-bus systems with doubly fed induction generator （DFIG） wind turbines demonstrate that， within a certain disturbance range， the proposed strategy can effectively raise the system frequency nadir， reduce rotor kinetic energy depletion， and alleviate the frequency regulation burden on conventional generators. The method provides a new solution for maintaining frequency stability in power systems with high penetration of renewable energy.
To address the decline in system inertia and frequency stability challenges associated with high penetration of renewable energy， this paper proposes a coordinated control method for hybrid energy storage systems （HESS） based on hierarchical power allocation and adaptive state-of-charge （SOC） management. The proposed strategy utilizes a three-layer hierarchical architecture： the top layer generates power commands by integrating inertial support and primary frequency regulation based on frequency deviation and its rate of change； the middle layer allocates transient power to supercapacitors and steady-state power to lithium-ion batteries， employing fuzzy logic control to prevent SOC limit violations； and the bottom layer utilizes an equalization algorithm to ensure SOC consistency among power units within each group. To validate the engineering applicability of the strategy， a hardware-in-the-loop （HIL） simulation platform was developed and tested. Results demonstrate that the proposed method effectively enhances system frequency stability with superior dynamic response and robust SOC management. By ensuring operational reliability and mitigating battery degradation， this study provides both theoretical and practical support for the deployment of energy storage power stations.
In application scenarios such as the interconnection of partitioned power grids where the input and output frequencies of the modular multilevel matrix converter （M3C） are identical， the coupling between the voltages and currents on both sides introduces an additional DC component into the arm power. This results in an arm power imbalance that must be mitigated. First， the generation mechanism and characteristics of this imbalance are analyzed， demonstrating that conventional balancing strategies based on fundamental-frequency circulating current injection become ineffective when the voltage magnitudes on both sides are comparable. Subsequently， a control strategy is proposed based on the injection of a third-harmonic arm common-mode voltage and circulating current. Subject to the constraint of not increasing the peak arm voltage， a method for selecting the optimal amplitude and phase of the third-harmonic common-mode voltage is derived. Furthermore， the relationship between the circulating current amplitude and the voltage transformation ratio and power factor is investigated. Finally， the effectiveness of the proposed strategy is validated through MATLAB/Simulink simulations.
In large-scale grid-connected systems of permanent magnet synchronous generator （PMSG） wind farms， grid-forming energy storage （GFMES） can provide virtual inertia to enhance grid stability； however， it also makes accurate assessment of small-signal stability more challenging. To address this issue， a stability index based on exponential input-to-state stability （EISS） theory is proposed. First， considering time-delay characteristics and the effect of virtual inertia， the PMSG wind farm grid-connected system considering GFMES is modeled as an equivalent time-varying discrete state-space model （TDSSM）. Second， based on the TDSSM， a global asymptotic gain matrix is derived to quantify the interactions among subsystems. Then， an exponential dynamic short-circuit ratio （EDSCR） is proposed based on EISS theory， and its necessary and sufficient conditions are rigorously established. The EDSCR is used to characterize the small-signal stability boundary of the PMSG-GFMES system. Finally， simulation studies and comparative analyses on the MATLAB platform verify the effectiveness of the proposed model and index.
To address the static voltage stability margin （SVSM） calculation problem in power systems with high penetration of renewable energy and emerging loads， this paper presents a probabilistic SVSM calculation method based on a multi-fidelity surrogate model that accounts for multi-source uncertainties from both generation and demand. The proposed multi-fidelity model consists of a low-fidelity model and a correction function. The low-fidelity model utilizes low-accuracy samples as input and leverages sparse polynomial chaos expansion to fully exploit the high computational efficiency of surrogate models. Concurrently， the correction function uses a small number of high-accuracy samples as input to modify the undetermined coefficients of key polynomial basis functions， thereby improving calculation accuracy while minimizing the computational burden. Finally， the validity of the proposed method is verified using the IEEE 30-bus and IEEE 118-bus systems， which incorporate wind and PV power plants and stochastic loads， and the impacts of model parameters are analyzed.
The oscillatory DC circuit breaker is an innovative DC circuit breaker technology that is expected to achieve fast interruption capability while reducing cost. To analyze and outline the development prospects of novel oscillatory DC circuit breaker technology， this paper first introduces the urgent demand for rapid fault isolation in flexible DC transmission systems and DC grids， and summarizes the cost and performance bottlenecks encountered by conventional DC circuit breaker technologies in engineering applications. Second， the topologies and operating principles of various novel oscillatory DC circuit breaker configurations， including active and passive oscillatory types， are systematically described and analyzed. A comparative analysis is conducted across multiple aspects， such as control complexity， number of devices， and conduction losses. Then， focusing on key components of oscillatory circuit breakers， the characteristics of power electronic switching devices and the arc characteristics of fast mechanical switches are analyzed. Finally， the challenges in oscillatory DC circuit breaker research are summarized， and future development directions are outlined， aiming to provide a reference for further research and engineering applications in this field.
To address the challenges in differential protection applications for multi-terminal lines in distribution networks， such as reduced sensitivity， the inability to compensate for capacitive current under no-voltage conditions， and susceptibility of current transformers （CTs） to saturation during external faults， a differential protection method based on the alpha plane of current fault components is proposed. This method builds on the alpha-plane protection principle， constructs an analysis region using vector information of current fault components， and discriminates between internal and external faults through dynamic partitioning. Theoretical analysis is conducted to evaluate the method’s immunity to capacitive current and CT saturation， and simulation studies are performed under various fault types， capacitive current levels， and CT saturation conditions. The results indicate that the proposed method features high sensitivity and a wide discrimination margin， is unaffected by load current， and does not require additional capacitive current compensation devices or CT saturation identification units， making it suitable for application in distribution networks.
With increasing uncertainty on both the generation and load sides in 220 kV supply areas， power fluctuations in transmission corridors have become more pronounced， leading to a growing risk of line overloading. To address this issue， a capacity optimization method for energy storage power stations considering flexible load transfer is proposed. First， the 220 kV power supply area is reasonably partitioned based on substation unit structures， providing a partitioning basis for modeling flexible load transfer. Then， a bi-level distributed robust optimization model is established. The upper level minimizes the investment cost of energy storage to determine optimal capacity and power ratings， while the lower level minimizes the coordinated operation cost of the 220~110 kV transmission and distribution system， considering power flow and network reconfiguration constraints. Source~load uncertainty is characterized using a composite norm， and the column-and-constraint generation （C&CG） algorithm is employed for efficient solution. Simulation results demonstrate that the proposed method effectively alleviates line overloading and enhances renewable energy accommodation， thereby verifying its effectiveness and engineering applicability.
